Research on Dynamic Scheduling of Grid Monitoring Data Processing Tasks Based Storm

Abstract

Development of smart grid spawned the big data in electric power industry, the cloud computing platform provided the solution for the big data in electric power industry, it has a significant effect for batch jobs, but its real-time is not guaranteed. For the real-time problem of cloud computing platform, the Storm platform will be introduced to monitor the grid power. This paper studies the fair share scheduling algorithm under the Storm platform. It introduced the concept of Storm framework briefly, then, proposed the fair share scheduling algorithm according to the lack of current Storm scheduling algorithm, finally, the experiment proved that the scheduling algorithm based on fair share improved the resource utilization of Storm cluster and reduced the processing delay of the data.